Like "things very rarely turn out like we expect",...
I thought it would be really funny to take the histogram of [list-wrandom] and send the results it creates back in as its own weights.
If you do this quickly and after fewer (than more) iterations variations occur which deviate from the original weights.
This abstractions also has a way to get back to the original "root" weights(reset), clear the histogram(clear), and "feedback" in the weights via list messages.
My thinking is...this will come in Really handy when one wants to markov chain randomness but "hold" a pattern by feeding back in the weights one likes. Sort of "Deviate from The Norm", then return to it.
If you look at it, I suggest you play around with the metro (in the help) and the metro times and try "clearing+return".
Enjoy.
Love through Sharing.
Peace through PD(!).
-s
p.s. I know it's a pd-thing, but I think this is really funny.